I think the FW3A is probably a more practical light than the D4. It has a clip, it has more 7135 chips for steady med/high modes, the beam should be rounder/smoother, and the maximum output might be a bit less intense.
I’m not sure the tail switch will be any better or worse — just different. It encourages a reverse grip instead of a forward grip.
Looking at djozz’s XP-G3 measurements, overall output might be pretty close to the D4, despite having 3 emitters instead of 4. It seems that XP-G3 puts out more lumens per amp than 219c, at similar Vf. It’s almost as bright as XP-L HI, but has lower Vf so it can run at higher amps. So I wouldn’t be surprised if the FW3A can hit 4000 lumens. This assumes all other components are as good as the D4 though, which might be difficult.
My main concern is that I haven’t really heard anything good about the XP-G3’s tint consistency. However, it might still be totally fine with a frosted optic like the 10511. I’d recommend this over the 10507 for almost any light, but particularly for one with LEDs which emit different tints at different angles.
If possible, it’d probably be a good idea to offer it in a range of NW tints from 4000K to 5000K, with high CRI for some options. That would cover the majority of people’s preferences. I’m still rather fond of high-CRI 4500K.
He wrote it’s from a smartphone. So I wouldn’t rely on correct color. But the light distribution is nice.
He wrote also about a light yellow ring on the outside typicall for xp-G3, but no artefacts.
NW is indeed the only option
No glue will be used so swapping emitters is possible, giving all the option to use preferred LEDs.
Since NW is more popular, I bet t will not be hard to trade the emitters with somebody else on the forum for the cost of postage
joechina is right, the picture was taken with my smartphone and should not be taken as reference, neither for color/bin/tint nor for CRI of the final LED. In fact I just took the image for explaining something during the conversion with LT and I used what I had next to me. I’ll try to make a proper beamshot in a week or so (have to check if I have a triple, normally I use LED’s with higher CRI) but in fact there won’t be much special on a possible beamshot. Until now we don’t intend to use any special/secret/new components and therefore the beamshot should look pretty much the same as every other beamshot with an XP-G3 and a Carclo 10507